Energy and Buildings 40 (2008) 486494 www.elsevier.

com/locate/enbuild

The cooling and heating potential of an earth tube system in buildings

Kwang Ho Lee *, Richard K. Strand
University of Illinois at Urbana-Champaign, Champaign, IL, United States Received 3 September 2006; received in revised form 3 April 2007; accepted 13 April 2007

Abstract A new module was developed for and implemented in the EnergyPlus program for the simulation of earth tubes. The model was validated against and showed good agreement with both theoretical and experimental data. Using the new module, a parametric analysis was carried out to investigate the effect of pipe radius, pipe length, air ow rate and pipe depth on the overall performance of the earth tube under various conditions during cooling season. Pipe length and pipe depth turned out to affect the overall cooling rate of the earth tube, while pipe radius and air ow rate mainly affect earth tube inlet temperature. The cooling and heating potential of earth tubes in four different locations were also investigated. Whether or not an earth tube is benecial turned out to be heavily dependent on the climate of the location. # 2007 Elsevier B.V. All rights reserved.
Keywords: Earth tube; EnergyPlus; Parametric analysis; Cooling and heating potential

1. Introduction The utilization of geothermal energy to reduce heating and cooling needs in buildings has received increasing attention during the last several years. An earth tube is a long, underground metal or plastic pipe through which air is drawn. As air travels through the pipe, it gives up or receives some of its heat to/from the surrounding soil and enters the room as conditioned air during the cooling and heating period. Due to the signicance of earth tube system, numerous research studies have been performed by Krarti et al. (1995) [1], Puri (1986) [5] and Labs at al. (1989) [2]. Krarti (1995) developed the model to determine the annual mean soil surface temperature and amplitude of soil surface temperature, which is essential to the calculation of soil surface temperature. Labs [2] developed the model to determine the soil temperature at various depths and elapsed time using the annual mean soil surface temperature and the amplitude of soil surface temperature. Recently, a sophisticated model describing the complex mechanisms of simultaneous heat and mass transfer occurring around the earth tube has been developed and integrated into TRNSYS by Mihalakakou et al. (1995) [4].

* Corresponding author at: 711 W. Main Street #6, Urbana, IL 61801, United States. Tel.: +1 217 419 6067. E-mail address: klee56@uiuc.edu (K.H. Lee). 0378-7788/$ see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2007.04.003

Nevertheless, those research studies have focused either on heat transfer to/from the surrounding soil or on the prediction of soil temperature separately. To date, a detailed algorithm calculating the soil temperature variation around the earth tube directly from weather data les has not been encoded within an existing simulation tools. Since an accurate ground temperature prediction is also an essential factor for the simulation of an earth tube, both the heat transfer occurring around the earth tube and the soil temperature should be modeled together in order to successfully simulate an earth tube. In order to compute the soil temperature surrounding the earth tube and the subsequent heat transfer rate, three important parameters should be determined: annual mean ground surface temperature, amplitude of the annual soil surface temperature variation and the phase constant of the soil surface. The existing algorithms require users to input these three values, but nding out these three values is another big challenge for users and it is done by experiments in most cases. In this circumstance, the new model is developed to calculate those three parameters directly from the weather data le in this study. The objective of this paper is to discuss the development and implementation of a new module handling both the heat transfer and soil temperature algorithms into the EnergyPlus program for the simulation of earth tubes. The model is validated against the data from other studies. Also, using the new model, the effects of four parameters on the overall performance of the earth tube is quantitatively assessed through the parametric analysis. In addition, the cooling and heating

K.H. Lee, R.K. Strand / Energy and Buildings 40 (2008) 486494

487

Nomenclature a As b Ca f hc hs kair kp ks L ma ra r1 r2 r3 Rc constant (103 Pa/8C) amplitude of the soil surface temperature variation (8C) constant (609 Pa) specic heat of air (J/kg 8C) fraction of evaporation rate convective heat transfer coefcient at the inner pipe surface (W/m2 8C) convective heat transfer coefcient at the soil surface (W/m2 8C) thermal conductivity of the air (W/m 8C) pipe thermal conductivity (W/m 8C) soil thermal conductivity (W/m 8C) pipe length (m) mass ow rate of ambient air through pipe (kg/s) relative humidity inner pipe radius (m) pipe thickness (m) distance between the pipe outer surface and undisturbed soil (m) thermal resistance due to convection heat transfer between the air in the pipe and the pipe inner surface (8C/W) thermal resistance due to conduction heat transfer between the pipe inner and outer surface (8C/W) thermal resistance due to conduction heat transfer between the pipe outer surface and undisturbed soil (8C/W) total thermal resistance between pipe air and soil (8C/W) radiation constant (63 W/m2) average solar radiation (W/m2) amplitude of the solar radiation (W/m2) time elapsed from beginning of calendar year (days) phase constant of the soil surface (s; days) phase constant of the air (s; days) air temperature above the ground surface (8C) air temperature of the pipe at the distance y from the pipe inlet (8C) ambient air temperature (8C) average soil surface temperature (8C) average air temperature (8C) ground surface temperature (8C) amplitude of the air temperature (8C) ground temperature at time t and depth z (8C) overall heat transfer coefcient of the whole earth tube system (W/8C) average pipe air velocity (m/s) annual angular frequency (=1.992 107 rad/s) depth of the radial center of pipe below soil surface (m)

Greek letters as soil thermal diffusivity (m2/s; m2/days) b soil absorption coefcient (=1 soil albedo) e hemispherical emittance of the ground surface wI phase angle between the insolation and the air temperature (rad) ws phase angle difference between the air and soil surface temperature (rad)

potential of the earth tube is investigated in four representative locations in the U.S. 2. Earth tube model description The simulation program in which the earth tube module was implemented is EnergyPlus, since EneryPlus is considered to be the next-generation building performance simulation program combining the best features of DOE-2 and BLAST. As a result, it has the great exibilities and capabilities for the comprehensive building simulation. The integrated solution manager in EnergyPlus consists of three managers: the surface heat balance manager, the air heat balance manager and the building systems simulation manager. The earth tube module presented in this paper was implemented at the air heat balance manager level. Due to the complex mechanisms occurring around the earth tube, several simplifying assumptions were made:  Convection ow inside the pipe is hydrodynamically and thermally developed.  Soil temperature in the pipe vicinity can be calculated using the soil model discussed below beyond a particular distance from the center of the pipe (thickness of the annulus).  The temperature prole in the pipe vicinity is not affected by the presence of the pipe. As a result, the pipe surface temperature is uniform in the axial direction.  The soil surrounding the pipe is homogeneous and has a constant thermal conductivity.  Pipe has a uniform cross sectional area in the axial direction. 2.1. Soil temperature calculation Prior to the calculation of the soil temperature around an earth tube, the ground surface temperature directly above earth tube should be predicted. According to Kusuda and Achenbach (1965) [8], the ground surface temperature satises the following expression: T sur T 0; t T m As Reeiwt (1)

Rp

Rs

Rt DR Sm Sv t t0 t0a Ta Ta( y) Tam Tm Tma Tsur Tva Tz,t Ut Va w z

where T(x,t) is the soil temperature prole as a function of depth x and time t. Tm and As are the annual mean value and the amplitude of the ground surface temperature variation, respectively, which should be calculated by considering the convective heat transfer between the air and ground, the solar radiation

488

K.H. Lee, R.K. Strand / Energy and Buildings 40 (2008) 486494

absorption by the ground, the long-wavelength radiation emitted from the soil, and the latent heat loss due to the moisture evaporation at the ground surface. The model is described in depth by Krarti et al. [1], and the derivation process will not be discussed in this paper. According to Krarti et al. [1], the annual mean ground surface temperature, Tm, can be estimated as follows: 1 T m hr T ma eDR bSm 0:0168hs fb1 r a he (2)

resistance values: Rc Rp Rs 1 2pr 1 Lhc 1 r1 r2 ln 2pLkp r1 1 r1 r2 r3 ln 2pLks r1 r2 (9) (10) (11)

where he hs 1 0:0168a f ; hr hs 1 0:0168ar a f More detailed description of the algorithm is provided in reference [7]. The amplitude of the soil surface temperature variation (8C), As, the phase constant of the soil surface (s), t0, and phase angle difference between the air and soil surface temperature (rad), ws, can also be determined as follows [1]: hr T va bSv eifI (3) As he dks t0 t0a fs w   hr T va bSv eifI fs Arg he dks (4) (5)

where Rc is thermal resistance due to convection heat transfer between the air in the pipe and the pipe inner surface (8C/W), Rp the thermal resistance due to conduction heat transfer between the pipe inner and outer surface (8C/W) and Rs is the thermal resistance due to conduction heat transfer between the pipe outer surface and the undisturbed soil (8C/W). The distance between the pipe outer surface and undisturbed soil, r3, is assumed to be equal to the radius of the pipe. The convective heat transfer coefcient at the inner pipe surface in Eq. (9) (W/m2 8C), hc, is a function of Nusselt number, Nu, and thermal conductivity of air (W/m 8C), which can be expressed by the following expression: hc Nu kair 2r 1 (12)

In Eqs. (3) and (5) the symbols jj jj and Arg are used to signify the modulus and the argument of a complex number, respectively. The phase constant of the air (s), t0a, is the time elapsed from the beginning of the year at which the air temperature reaches the minimum value in the year. The value of d is evaluated as follows: d 1i D (6)

Using the three thermal resistance values, Rc, Rp and Rs, overall heat transfer coefcient of earth tube can be estimated as follows: Ut 1 Rt (13) (14)

Rt Rc R p R s

where the dampening depth (m), D, is calculated from the following equation: r 2a s D (7) w Assuming homogeneous soil of constant thermal diffusivity, the temperature at any depth z and time t can be nally estimated by the following expression (Labs et al. 1989).   1=2  p T z;t T m As exp z 365as      2p z 365 1=2 cos t t0 (8) 365 2 pas 2.2. Heat transfer and earth tube inlet air temperature calculation In order to calculate the heat transfer between the earth tube and the surrounding soil, the overall heat transfer coefcient should be determined using the following three thermal

Now, the heat transfer between the soil and the air inside the pipe is equal to the amount of heat loss or gain as air ows along the pipe [3]: U t dyT a y T z;t m a C a dT a y (15)

By solving for air temperature inside the pipe, Ta( y), the following earth tube inlet air temperature (dened as the air leaving the earth tube and entering the space in this paper) can be nally obtained.  In case Tam > Tz,t T a L T z;t eA  In case Tam = Tz,t T a L T z;t  In case Tam < Tz,t T a L T z;t eA where A m a Ca lnjT am T z;t j U t L m a Ca (19) (16) (17)

(18)

K.H. Lee, R.K. Strand / Energy and Buildings 40 (2008) 486494

489

3. Model validation The model developed in the previous section was validated against both theoretical model and experimental data. The theoretical model for the comparison was developed by AlAjmi et al. [3], which was validated against relevant experimental and theoretical studies. The experimental study was carried out by Dhaliwal et al. [6] at North Carolina under the conguration of a pipe diameter of 30 cm, a pipe length of 24.7 m, and a pipe depth of 1.7 m. The detailed input parameters for the comparison to both experimental and theoretical studies are described in Table 1. The results for both the experimental and theoretical studies using two different ambient air temperature conditions are presented in Tables 2 and 3, Figs. 1 and 2. As can be seen, the results show good agreement and have the values between the two studies under consideration. The difference in the earth tube inlet temperature between the experimental data and the modeling results can be seen to be a result of an obvious data error in the experimental results (see Fig. 1). However, the discrepancies between the EnergyPlus model and the Al-Ajmi model are insignicant, indicating that the model can properly predict the performance
Table 1 Input parameters for comparative validation Pipe diameter (cm) Pipe length (m) Air velocity (m/s) Soil temperature (8C) Pipe depth (m) Soil thermal conductivity (W/m8C) Soil thermal diffusivity (m2/s) 30 24.7 1.5 18.89 2.13 1.16 0.00232

Fig. 1. E+ earth tube model validation (ambient air temperature 25.56 8C).

of an earth tube system and the heating/cooling load reduction due to the earth tube. 4. Description of simulation conditions and example building Using the newly developed earth tube model, the cooling and heating potential of an earth tube system was predicted in four different locations which represent four typical climatic conditions in order to investigate the inuence of soil temperature and soil condition. Also parametric studies were carried out to determine the effect of four important variables inuencing the earth tube inlet air temperature: pipe radius, pipe length, air ow rate and pipe depth under the ground surface. Simulations were performed on ve different values of each parameter while the other parameters were maintained at the same values. 21 August is chosen as summer design day for Key West and Spokane, and 21 July was chosen for Peoria and Phoenix. The maximum dry bulb temperatures were set at 30.6 8C, 30.4 8C, 35.7 8C and 28.3 8C for Key West, Peoria, Phoenix and Spokane, respectively. 21 December was chosen as winter design day for Phoenix and Spokane, and 21 January was chosen for Key West and Peoria. The maximum dry bulb

Table 2 E+ earth tube model validation (ambient air temperature: 25.56 8C) Axial distance from the pipe inlet (m) 3.35 6.4 9.45 12.5 15.55 24.7 Experimental data of Dhaliwal et al. (8C) 25.00 24.40 25.00 24.40 23.80 23.80 Theoretical data of Al-Ajmi et al. (8C) 24.94 24.43 23.97 23.54 23.15 22.16 E+ earth tube model (8C) 25.04 24.60 24.19 23.82 23.46 22.55

Table 3 E+ earth tube model validation (ambient air temperature: 20.55 8C) Axial distance from the pipe inlet (m) 3.35 6.4 9.45 12.5 15.55 24.7 Experimental data of Dhaliwal et al. (8C) 20.55 20.00 20.00 20.00 20.00 20.00 Theoretical data of Al-Ajmi et al. (8C) 20.40 20.27 20.16 20.05 19.95 19.71 E+ earth tube model (8C) 20.42 20.31 20.21 20.11 20.03 19.80

Fig. 2. E+ earth tube model validation (ambient air temperature 20.55 8C).

490 Table 4 Description of simulation conditions Conditions Location Spokane, WA: mild and dry Peoria, IL: mild and wet Phoenix, AZ: hot and dry Key West, FL: hot and wet Run period Summer design day 7/21: Peoria and Phoenix 8/21: Key West and Spokane Winter design day 12/21: Phoenix and Spokane 1/21: Key West and Peoria

K.H. Lee, R.K. Strand / Energy and Buildings 40 (2008) 486494

5. Parametric analysis The effects of four important parameters signicantly affecting the performance of an earth tube system were investigated through the parametric analysis. Since the ndings discussed later in this paper showed that earth tubes have more potential for cooling than for heating, the parametric analysis was carried out only for cooling. 5.1. Inuence of pipe length Fig. 3 presents the effect of pipe length on the earth tube inlet air temperature at the highest ambient air temperature. As the pipe length increases, the inlet air temperature decreases due to the fact that the longer pipe provides a longer path over which heat transfer between the pipe and the surrounding soil can take place given the same overall heat transfer coefcient of earth tube. However, the temperature range and decrease rate in terms of pipe length were different among each location due to the different soil conditions, ambient air temperature, and soil temperature distribution. In addition, Peoria and Spokane are appropriate locations to employ earth tubes since the earth tube inlet air temperature is lower than the specied indoor temperature of 23 8C. In both cases of Peoria and Spokane after a certain point around 5070 m increasing the length does not result in much better performance and the improvements begin to level off, indicating that these values can be the optimal design values in these special cases which can be determined using the developed earth tube model. 5.2. Inuence of pipe depth Fig. 4 shows the inuence of pipe depth under the ground surface on the earth tube inlet air temperature. As the pipe depth increases, the inlet air temperature decreases, indicating that the earth tube should be placed as deeply as possible. However, the trenching cost and other economic factors should be considered when installing earth tubes. Similarly the temperature range and decrease rate with regard to pipe depth were different at each location due to

Variables Pipe radius: 0.05 m, 0.075 m, 0.1 m, 0.15 m, 0.2 m Pipe length: 10 m, 30 m, 50 m, 7 0m, 90 m Air velocity: 2 m/s, 5 m/s, 8 m/s, 11 m/s, 14 m/s Pipe depth: 1 m, 2.5 m, 4 m, 5.5 m, 7 m

temperatures were set at 14.3 8C, 17.7 8C, 4.5 8C and 12.3 8C for Key West, Peoria, Phoenix and Spokane, respectively. Table 4 shows the details of the parametric studies. The base case values for each variable were set at: 0.075 m for pipe radius, 30 m for pipe length, 5 m/s for air velocity, and 2.5 m for pipe depth. These values were selected for the prediction of cooling and heating potential of the system. In addition, when changing only one variable at every simulation process for parametric studies, the other variables were kept at those values. Table 5 describes the inputted soil conditions and parameters for each location. The annual average ground surface temperature, Tm, and amplitude of the soil surface temperature variation, As, were calculated by a utility program that is provided with EnergyPlus. A three-zone residential building was chosen for this study. The building is not actually existing building, but the constructions and thermal conditions are thoroughly based on the real buildings. It consists of a living space, an attached garage and attic above living space and garage having oor areas of approximately 140 m2, 37 m2, and 177 m2, respectively. In this study, the living space will be analyzed, which is located on the northern side of the building with the ceiling height of 3.05 m. The internal heat gains for lights and equipment were set at 5.4 W/m2 and two people were placed in the living space during the simulation Inltration was set at 0.25 ACH and the earth tube was set to run constantly at the same volumetric ow rate of 285 m3/h during the whole running period.
Table 5 Soil related parameters Soil condition Key West Peoria Phoenix Spokane Heavy Heavy Heavy Heavy and and and and moist moist dry dry Tm (average) (8C) 24.3 10.0 25.0 9.9 As (amplitude) (8C) 5.0 18.8 9.4 18.5

Fig. 3. Inuence of pipe length on inlet temperature.

K.H. Lee, R.K. Strand / Energy and Buildings 40 (2008) 486494

491

Fig. 4. Inuence of pipe depth on inlet temperature.

Fig. 6. Inuence of pipe radius on inlet temperature.

different soil and weather conditions. In addition like the case of pipe depth, Peoria and Spokane are the better places for the utilization of earth tubes than Key West and Phoenix which have higher soil temperatures and the earth tube inlet temperature. Moreover, after a certain point (around 4 5.5 m) the improvements begin to level off, which can be considered to be the optimal design values without factoring in economic factors such as trenching and drilling costs. Based on these results, pipe depth appears to have as large of an inuence on earth tube performance as pipe length. 6. Inuence of air velocity inside pipe Fig. 5 presents the effect of air velocity inside the pipe on the earth tube inlet air temperature. As the air ow rate increases the inlet air temperature increases in all locations, since the air spends more less in the tube and thus in contact with the lower soil temperature. This can be seen in the earth tube modeling equations since according to Eq. (19) a higher air ow rate causes a higher mass ow rate and higher air temperature. Likewise, the range and rate of increase of the inlet air temperatures as a function of air velocity were different at each location due to different soil conditions. Like the cases of other parameters described above, Peoria and Spokane turned out to be more appropriate locations for earth tubes, and there is a air

velocity limit above after which the inlet temperatures become relatively constant despite additional increases in velocity. However, when considering the air ow rate during the design process, simply reducing the ow rate does not necessarily improve the earth tube performance since the cooling heat transfer rate due to the earth tubes depends on both air ow rate and temperature difference, not on each factor alone (q = maCaDT). Thus, both the air ow rate and the temperature difference should be considered simultaneously. 6.1. Inuence of pipe radius Fig. 6 illustrates the effect of pipe radius on the earth tube inlet air temperature. As the pipe radius increases, the earth tube inlet air temperature also increases due to the fact that higher pipe radius results in a lower convective heat transfer coefcient on the pipe inner surface and a lower overall heat transfer coefcient of earth tube system. Similarly, the temperature range and the increase in the inlet temperature in terms of pipe radius were different at each location. Peoria and Spokane can be considered to be better places for earth tubes due to a lower inlet air temperature than the specied indoor temperature of 23 8C. Compared to the other three parameters discussed above, pipe radius did not affect the results as much as the other parameters. Moreover, simply reducing the pipe radius under same air ow rate will increase the air velocity inside the pipe, resulting in an increase in the earth tube inlet air temperature. Thus pipe radius and air ow rate should be considered together and optimized using simulated results. The trends of the results in terms of the inuence of design parameters on the performance discussed above were similar to those of other studies (Mihalakakou et al., 1995). 7. Cooling and heating potential The simulations were carried out for the investigation of the cooling and heating potential of an earth tube by the comparison of the results with and without earth tubes. Figs. 710 illustrate the cooling and heating rate variation under design day conditions with and without an earth tube in four

Fig. 5. Inuence of air velocity on inlet temperature.

492

K.H. Lee, R.K. Strand / Energy and Buildings 40 (2008) 486494

Fig. 7. Cooling/heating rate with and without earth tube (Key West).

Fig. 10. Cooling/heating rate with and without earth tube (Spokane).

Fig. 8. Cooling/heating rate with and without earth tube (Peoria).

different locations. Table 6 describes the required cooling load requirement of the building to maintain the indoor air temperature of 23 8C during the whole summer design day period with and without the earth tube. Similarly Table 7 describes the required heating load requirement of the building to maintain the indoor air temperature of 20 8C during the winter design day with and without the earth tube. As can be

seen in Figs. 710, the cooling rate varies more than the heating rate in all four locations since winter design days have constant ambient air temperatures. The increases in the cooling rate around hours 1220 are due to the increase in the ambient air temperature during that time period. The cooling rate around 2 8 h reaches zero in Peoria and Spokane due to the fact that ambient air temperature is low enough to maintain the comfortable indoor temperature without air-conditioning and earth tubes. In case of cooling performance, whether or not the earth tube is benecial clearly depends on climatic conditions and locations as can be seen in Table 6. The amount of reduced total cooling load requirement due to an earth tube system were 6.6 kWh and 8.8 kWh in case of Peoria and Spokane, respectively, an energy saving of 31% and 52%, respectively. Therefore, large amounts of cooling energy can be saved by employing earth tube systems. However, cooling load requirements with an earth tube system is even 41.8 kWh and 26.9 kWh higher than those without an earth tube in Key West and Phoenix, respectively. This is due to the fact that the earth tube inlet air temperatures were higher than specied indoor temperature in the two locations, indicating that the hot weather
Table 6 Total cooling load requirement of the building with and without an earth tube Locations Key West Peoria Phoenix Spokane Cooling load requirement without earth tube (kWh) 88.3 20.8 79.7 16.9 Cooling load requirement with earth tube (kWh) 129.7 14.3 106.6 8.1

Table 7 Heating load requirement of the building with and without an earth tube Locations Key West Peoria Phoenix Spokane Heating load requirement without earth tube (kWh) 20.9 170.7 64.7 142.7 Heating load requirement with earth tube (kWh) 21.2 196.1 65.7 161.3

Fig. 9. Cooling/heating rate with and without earth tube (Phoenix).

K.H. Lee, R.K. Strand / Energy and Buildings 40 (2008) 486494 Table 8 Total cooling load requirement under different earth tube air ow rates (design day simulation) Conditions Cooling load requirement without earth tube (kWh) Cooling load requirement with earth tube (0.75 ACH) (kWh) Cooling load requirement with earth tube (1.5 ACH) (kWh) Cooling load requirement with earth tube (3.0 ACH) (kWh) Peoria 20.9 14.3 12.8 14.2 Spokane 16.9 8.1 6.2 7.0 Key West Peoria Phoenix Spokane

493

Table 10 Heating load reduction by preheating the fresh air using an earth tube (design day simulation) Locations Heating load requirement with 0.35 ACH inltration rate (kWh) 22.3 184.1 69.7 153.3 Heating load requirement with 0.35 ACH earth tube air ow rate (kWh) 14.5 159.3 52.0 132.3

of the two locations also increased the soil temperature signicantly and that the weather and soil conditions should be considered in order to determine whether or not earth tubes should be used in particular locations. Furthermore, additional simulations were carried out to investigate the cooling load reduction variation under different earth tube air ow rate conditions: 0.75, 1.5 and 3.0 ACH. Since Peoria and Spokane showed the cooling load reduction due to an earth tube, simulations were carried out only for these two locations. Tables 8 and 9 show the total cooling load requirements under different earth tube air ow rates simulated during the design day and the whole cooling season respectively. As can be seen in the tables, large portions of the cooling loads are reduced due to earth tubes. In case of the whole cooling season simulation, more than 50% of cooling loads can be saved in both locations. In both the design day and the whole cooling season simulations, the earth tube air ow rate of 1.5 ACH turned out to cause the largest total cooling load reduction, even larger than 3.0 ACH. This is due to the fact that the earth tube inlet air temperature becomes higher as the air velocity inside the pipe also gets higher due to the higher earth tube air ow rate under the same pipe radius condition, which can reduce the cooling provided by the earth tube to the zone. Therefore, it should be noted that the higher earth tube air ow rate does not necessarily mean the highest cooling load reduction due to the earth tube, indicating that the optimal air ow rate under different conditions should be determined in advance using an earth tube model in a program such as EnergyPlus. Although the earth tube system did not appear to have the capacity to replace the conventional air-conditioning
Table 9 Total cooling load requirement under different earth tube air ow rates (June August simulation) Conditions Cooling load requirement without earth tube (kWh) Cooling load requirement with earth tube (0.75 ACH) (kWh) Cooling load requirement with earth tube (1.5 ACH) (kWh) Cooling load requirement with earth tube (3.0 ACH) (kWh) Peoria 1099.8 427.1 336.5 389.5 Spokane 1147.8 497.2 394.4 444.5

system completely, it can signicantly reduce the cooling load requirement to maintain the specied indoor conditions when the weather and soil conditions are properly considered for each location. On the other hand, in case of heating performance, the heating load requirements in all the four locations with an earth tube were higher than those without an earth tube as can be seen in Table 7. This is due to the fact that the earth tube inlet air temperatures are not higher than the specied indoor temperature within the comfort zone, indicating that an earth tube inlet temperature high enough to provide heating cannot be obtained. Therefore, an earth tube system has better capability for cooling than for heating buildings given the same location. Although an earth tube did not turned out to reduce the heating load, additional simulations were carried out to nd out the contribution of earth tubes to the preheating of the fresh air. According to ASHRAE Standard 62-1999, at least 0.35 ACH of the air should be ventilated for living spaces. Therefore, two cases were simulated: the case with the inltration rate of 0.35 ACH without the assistance of an earth tube, and the case with the earth tube air ow rate of 0.35 ACH. Tables 10 and 11 show the heating load requirements of the two cases simulated during the design day and the whole heating season, respectively. As can be seen in the tables, a considerable amount of the heating load can be reduced through the preheating of the fresh air using earth tubes, though not as much as the cooling loads. In the case of Phoenix simulated for the whole heating season, 45% of the heating loads can be saved by preheating the fresh air. The reason why the heating loads in Key West simulated for the whole heating season are zero is that most ambient air temperatures were higher than the specied indoor air temperature of 20 8C even in the winter season and that the indoor temperature did not fall below the 20 8C at any point during the simulation.

Table 11 Heating load reduction by preheating the fresh air using an earth tube (DecemberFebruary simulation) Locations Heating load requirement with 0.35 ACH inltration rate (kWh) 0 9513.6 1788.0 8283.9 Heating load requirement with 0.35 ACH earth tube air ow rate (kWh) 0 8542.6 990.1 7564.3

Key West Peoria Phoenix Spokane

494

K.H. Lee, R.K. Strand / Energy and Buildings 40 (2008) 486494

8. Conclusion In this paper, the algorithm for the simulation of earth tubes is described, and validated against the data of other studies. Parametric studies were carried out to investigate the effect of each parameter on earth tube. The cooling and heating potential of an earth tube was also investigated using the newly developed model, and the following conclusions can be made based on the data presented above: 1. As a result of the validation against the data from other theoretical and experimental studies, the earth tube model showed a good agreement with the work performed by others. Thus it can be suitably used to predict the thermal performance of earth tube systems. 2. A deeply placed and longer earth tube with a lower air velocity and smaller radius should result in lower earth tube inlet temperatures. However, the trenching cost and other factors should also be considered when installing earth tubes. 3. Pipe length, air velocity inside pipe and pipe depth turned out to have more inuence on earth tube performance than pipe radius. However, pipe radius and air ow rate as well as cooling heat transfer rate should be considered simultaneously. 4. When properly designed, an earth tube were shown to save more than 50% of the total cooling load in the cases presented in this paper, depending on the weather and soil conditions. Although the earth tube alone cannot replace conventional air-conditioning system in these case studies, it can signicantly reduce the cooling load in buildings. 5. Earth tubes turned out to have a larger capacity for cooling than for heating given the same location. Even among the case of cooling, whether or not earth tubes are benecial is dependent on the local climate, indicating that weather and soil conditions should be properly considered and investigated using the new model.

The availability of an earth tube model in a program such as EnergyPlus is an important step forward when attempting to determine whether or not earth tubes should be used for a particular building and to determine the most optimal combination with regard to depth, length, radius, and air velocity. Based on this study, future work that should be done includes the economic assessment and the environmental effects of the earth tube since those factors affect the feasibility of the system. Acknowledgment The authors of this paper wish to thank the U.S. Department of Energys Lawrence Berkeley National Laboratory for funding the work which led to this paper under Grant 6712530. References
[1] M. Krarti, C. Lopez-Alonzo, D.E. Claridge, J.F. Kreider, Analytical model to predict annual soil surface temperature variation, Journal of Solar Energy Engineering 117 (1995) 9199. [2] K. Labs, in: J. Cook (Ed.), Passive Cooling, MIT Press, Cambridge, Massachusetts, 1989. [3] F. Al-Ajmi, D.L. Loveday, V.I. Hanby, The cooling potential of earth-air heat exchangers for domestic buildings in a desert climate, Building and Environment 41 (2005) 235244. [4] C.P. Jacovides, G. Mihalakakou, An underground pipe systems as an energy source for cooling/heating purposes, Renewable Energy 6 (1995) 893900. [5] V.M. Puri, Heat and Mass Transfer Analysis and Modeling in Unsaturated Ground Soils for Buried Tube Systems, Energy in Agriculture 6 (1987) 179193. [6] Dhaliwal, Goswami, Heat Transfer Analysis Environment Control using an Underground Air Tunnel, ASME Solar Energy Div., Las Vegas, 1984, pp. 505510. [7] K.H. Lee, R.K. Strand, Implementation of an earth tube system into EnergyPlus Program, in: Proceedings of the SimBuild 2006 Conference, Boston MA, USA, 2006. [8] T. Kusuda, P.R. Achenbach, Earth Temperature and Thermal diffusivity on at Selected Stations in the United Stales, ASHRAE Trans. 71 (1965) 6175.